The characterization of cellular gene expression finds application in a variety of disciplines, such as in the analysis of differential expression between different tissue types, different stages of cellular growth or between normal and diseased states. Fundamental to differential expression analysis is the detection of different mRNA species in a test population, and at least the qualitative, if not quantitative, determination of different mRNA levels in that test population.
In many currently employed gene expression analysis protocols, detection of different mRNA levels involves the steps of generating an image or target nucleic acid population that is representative of the mRNA population of the test sample. In other words, a population of image nucleic acids is generated where the population is indicative of the different mRNAs that are originally present in the sample. The image may be DNA or RNA and may have the sequence of initial mRNA or the complement thereof. Following generation, the population of image or target nucleic acids is hybridized to an array of probe nucleic acids stably associated with the surface of a solid support. Since the sequence and location of each probe is known, any resultant hybridization complexes that form on the array surface between target and probe can be used to identify those genes that are expressed in the cell from which the initial mRNA sample was obtained. Since the methods require detection of target/probe complexes on the array surface, the target nucleic acids are generally labeled so that they can be detected. Of increasing interest is the use of fluorescent labels for target nucleic acid labeling.
There are generally two ways of labeling image or target nucleic acids with fluorescent labels--direct fluorescent labeling protocols and indirect fluorescent labeling protocols. In the case of direct fluorescent labeling, a fluorescently labeled chemical analog of one or more of the four nucleoside triphosphates (henceforth called nucleotides) is present during the enzymatic generation of the image or target nucleic acid, which results in incorporation of the labeled nucleotide into the polymerized nucleic acid and consequent direct fluorescent labeling of the nucleic acid. While direct fluorescent labeling is employed in many protocols, there are disadvantages associated with this approach. First, not all available fluorophores are amenable to incorporation into the image polynucleotide during its generation, i.e. not all fluorophore-labeled nucleotides are efficiently incorporated by nucleic acid polymerases, which limits the availability of useable labels. Second, fluorescently labeled nucleotides that are capable of being processed by nucleic acid polymerases are not processed efficiently, resulting in significant waste of reagents and costs associated therewith. In addition, the fluorescently labeled nucleotides are often expensive.
In order to address at least some of the above disadvantages associated with direct fluorescence labeling techniques, indirect fluorescent labeling protocols have been developed. In one type of indirect fluorescent labeling protocol, nucleotides modified with chemical tags, e.g. biotin, are employed in the image nucleic acid generation step. Next, either during or after hybridization, a fluorescent label modified to specifically bind to the chemical tag, e.g. streptavidin conjugated fluorescent label, is contacted with the image or target nucleic acid, thereby providing indirect fluorescent labeling. While indirect labeling protocols such as those described above overcome some of the disadvantages of direct labeling protocols, they are not entirely satisfactory. For example, such protocols are not readily adaptable to "two color" gene expression analyses, in which two different populations of target nucleic acids are hybridized to the same array and distinguished from each other by means of distinguishable labels.
Accordingly, there is interest in the development of improved methods of indirectly fluorescently labeling image or target nucleic acids for use in gene expression analyses. Of particular interest would be the development of an indirect labeling protocol suited for use with two color hybridization procedures in which mixed populations of two or more different image polynucleotides which do not include chemically modified nucleotide residues can be labeled with distinguishable fluorescent labels.
Relevant Literature
U.S. patents of interest include: U.S. Pat. Nos. 5,604,097; 5,635,400; 5,695,934; 5,763,175; 5,863,722; and 5,846,719. Also of interest are: WO 97/31256 and WO 98/24933.